2023 - A year in review - Navigating the Era of Large Language Models

It's been a while since I wrote a "year in review" post. It's been a long time since I published anything actually. While I still love to write, read and research, I find that actual publication has become increasingly challenging. For me, that is mostly because of the ascent of Large Language Models (LLMs) in 2023.

Large Language Models

Fascination with LLMs

ChatGPT continues to amaze me. Since its release, I've explored its capabilities and remain baffled by what it can do. Especially if you take into account that it utilitises advanced pattern matching rather than true comprehension. It's a truly remarkable technology.

I firmly believe that LLMs are going to be a disruptive force. Their impact on both our professional and personal lives will be profound. The ability to extract information through prompt engineering instead of traditional linear learning will open up endless possibilities. It will transform the way we learn and interact with vast repositories of knowledge, speeding up the process from days to minutes.

We can now outsource menial jobs to Copilot. It can generate presentations, meeting notes, summaries, code and much, much more. In the development and digital workspace, the potential is limitless.

The Challenge Ahead

Any new technology of this scale is bound to have challenges. History has shown that the promises of technological advancements often lead to unintended consequences. Social media, the internet, and the industrial revolution all fell short of their utopian visions. Will LLMs follow suit? Are we ready for a technological change of this magnitude?

It's strange to witness the reverse effects of promises that come with game changing technological innovations. Social media, designed to facilitate connected and meaningful online lives, has only left us lonely and with compromised authenticity, and has become a breeding ground for hate and racism. The internet, a tool which would democratize knowledge, leaves us questioning its actual impact on children's education and development in its current form. And similarly, the industrial revolution's pledge of increased leisure time resulted in more busy lives and larger profits for large corporations, accompanied by disastrous environmental consequences and climate change.

As we stand on the brink of the rise of (generative) AI, scepticism persists. Will this new technology follow the pattern of its predecessors, delivering unintended, adverse and possibly grave outcomes? "When the wrong person uses the right means, the right means work in the wrong way". This should be a cautionary reminder, urging us to approach the integration of AI with careful consideration and ethical foresight.

Dumbification by Generative AI

A significant concern is the potential "dumbification" caused by outsourcing essential tasks to AI. If we entrust activities like research, writing, and creativity to AI, do we risk losing our own capabilities in these areas? I love to write, but I have got to acknowledge the fact that ChatGPT does it better. It's really disheartening to see a tool rewrite my posts and ending up with a better result. So this begs the question, why do these things if a trained AI model does it better?

Generative AI poses a genuine threat to the identity of humankind. "Who we are" as a species is answered by "what we do". We are our art, we are our music, we are our books. Our love for these pursuits, the whole of human creativity, defines who we are as a species. Overcoming the temptation of AI's proficiency will be crucial in the years to come.

A year of -ing instead of coding

Evolution of Roles

In 2023, my focus shifted from coding to various "-ing" activities such as architecting, photographing, and traveling. As a Lead Architect, my role evolved, but the essence of problem-solving remained constant. The tools changed, but the purpose endured.
Looking ahead to 2024, I know one thing for sure. Coding is firmly in my past, and I can't wait to see what sort of "-inging" it will replace.

What I was into this year

  • I have started using Obsidian as my main note taking app. The concept of having a second brain is something I really like and am still experimenting with.
  • I have used Llama to play around with Generative AI.
  • I moved from Pocket to Omnivore. Which has been better in every aspect. Next year I'll look into Readwise which is superior, but also comes with a price tag.
  • I started taking photography seriously, took evening classes, and that will likely continue in 2024. I'm now rocking a Canon R7 which upgraded my old-ish Canon M50
  • I read Creativity, Inc., Extreme Ownership and The Precipice.
  • I bought an BOOX Tab Ultra C to replace my Remarkable (of which I absolutely hate the closed eco system).

Conclusion

2023 was a year of technological exploration and personal evolution. While the rise of LLMs introduces challenges, as the gatekeepers of technology, we will have to find ways to safeguard our collective core identity. As we step into the new year, may it be filled with innovation, self-discovery, and a collective commitment to navigate this continouosly evolving landscape. I wish you all the very best!

Photo by Kajetan Sumila on Unsplash